Page 4 of 5

Re: Porteus-v2.0-i486 bug reports

Posted: 19 Jun 2013, 14:17
by fanthom
arch independent and will work for every release:

Code: Select all

#!/bin/bash
# convert Slackware's TGZ (or TXZ) package into .xzm compressed file
# which can be used as a LiveCD module
#
# Author: Tomas M. <http://www.linux-live.org>
# Modifications for Porteus by fanthom

mod=`readlink -f $1`
mod2=`readlink -f $2`
# Switch to root
if [ "$DISPLAY" ]; then
   if [ `whoami` != "root" ]; then
      ktsuss "/opt/porteus-scripts/txz2xzm $mod $mod2"
      exit
   fi
else
   if [ `whoami` != "root" ]; then
      echo "Please enter root's password below"
      su - -c "/opt/porteus-scripts/txz2xzm $mod $mod2"
      exit
   fi
fi

if [ "$1" = "" -o "$2" = "" ]; then
   echo
   echo "Convert Slackware's TXZ package into .xzm compressed module"
   echo "usage: $0 source_filename.txz output_file.xzm"
   exit 1
fi

TMPDIR=/tmp/txz2xzm$$

installpkg -root $TMPDIR $1
if [ $? != 0 ]; then echo "error installing package"; exit; fi

# optimalization procedures, this doesn't hurt
find $TMPDIR/usr{/local/,/share/,/}{man,info} -type l -name "*.gz" 2>/dev/null | xargs -r gunzip -f
find $TMPDIR/usr{/local/,/share/,/}{man,info} -type f -name "*.gz" 2>/dev/null | xargs -r gunzip
cp -a $TMPDIR/usr/share/man $TMPDIR/usr 2>/dev/null; rm -rf $TMPDIR/usr/share/man
rm -f $TMPDIR/{usr,usr/local,var}/man/cat*/*

dir2xzm $TMPDIR "$2"
if [ $? != 0 ]; then echo "error building compressed image"; exit; fi

rm -Rf $TMPDIR

Re: Porteus-v2.0-i486 bug reports

Posted: 19 Jun 2013, 16:02
by donald
not a bug just an improvement

as I am only using the always-fresh mode for browsing
I configured everything to my liking and let Porteus built
the -save your current session- changes module

next boot into always-fresh and using this changes-module
shows that some settings were not restored

-window-theme was daloa, fallback to evolve
and also all of my panel-settings are lost.It is the pure default one

I can live with that but it is a little boring to reconfigure
it again and again and again

as I said not really a bug....just for comfort
I'm using 32bit-XFCE

Re: Porteus-v2.0-i486 bug reports

Posted: 20 Jun 2013, 02:05
by Rava
Thanks fanthom for that txz2xzm...
donald wrote:not a bug just an improvement
I can live with that but it is a little boring to reconfigure
it again and again and again

as I said not really a bug....just for comfort
I'm using 32bit-XFCE
I ran into the same issue when using Porteus from 1.0 to 1.2 and even 2.0. both i486 or x86-64 versions.

Even when I just exit x and restart it (I use runlevel 3 as standard runlevel and start X from the virtual console by hand) XFCe remembers som4e settings, but always forgets some others.

As it is indeed bothersome to reconfigure stuff again and again, I looked into the config files and copied these by hand into rootcopy.

Still strange how the changes could be stored in the users XFCe folders and files and still would partially ignored when restarting X...

Mouse / Keyboard bug

Posted: 04 Jul 2013, 03:05
by Rava
At times the mouse (external, USB) stops responding in X, it still works well in the virtual console, though, and only after I restarted X all is well again.
But I always lose all my current work cause I cannot save any documents in all of the running X programs.

This never happened in any of my other Linux. Slax or Porteus versions, only when running Port 2.0 i486, on that one machine. (Samsung Sub-Notebook "Q40 series")

I get this error on the virtual console I manually started X from, when mousepad was the last foreground program:

Code: Select all

(mousepad:17629): Gtk-CRITICAL **: IA__gtk_widget_event: assertion `WIDGET_REALIZED_FOR_EVENT (widget, event)' failed
Trying to change the program to save some of my work via keyboard also only results in the same error as above. Also, desktop changing, e.g. via Ctrl+F1 or such no longer works as well...

I get a similar Gtk-CRITICAL error when any other program was the last foreground program. Just the beginning part (program-name:PID) differs.

What can I do to investigate this bug further?

I can still move the mouse in X, but non of the mouse button do anything at all, so I always loose my most current work.
For now killing X via virtual consoole is the only working solution. Unplugging the mouse and re-plugging also fails to help.

killall -9 mousepad
is not able to kill the processes as well...

The only thing that works is going into the virtual console that started X and kill X via Ctrl+C.

(Seems the only stuff that still works in X is changing to the virtual console)

_______________________________________________________________

Again, it happened.
Again, mousepad was the topmost X program.

Code: Select all

(mousepad:17982): Gtk-CRITICAL **: IA__gtk_widget_event: assertion `WIDGET_REALIZED_FOR_EVENT (widget, event)' failed
Again, nothing else works in X but changing to the virtual console. I was editing some stuff in mousepad but are not able to save the current changes, nor able to continue.
All I could do is change to X (writing this stuff in the virtual console...) and taking a real screenshot with a real camera... to at least have the most recent work "saved" on the last viewed X workspace, since I cannot change programs or workspaces anymore.

Any ideas at all how to investigate further when it again happens?

Re: Porteus-v2.0-i486 bug reports

Posted: 04 Jul 2013, 05:46
by fanthom
sounds like a hardware/kernel related but please test KDE4 desktop (just in case) and check if the problems persist.

Re: Porteus-v2.0-i486 bug reports

Posted: 04 Jul 2013, 13:49
by brokenman
I will also test here with my USB mouse.

Re: Porteus-v2.0-i486 bug reports

Posted: 05 Jul 2013, 01:58
by Rava
fanthom wrote:sounds like a hardware/kernel related but please test KDE4 desktop (just in case) and check if the problems persist.
How can I setup this, using my erxisting modules, but replacing rootcopy and the xfce modules?

I know that I can define a certain different path to rootcopy, but can I exclude and include certain modules as well?

Re: Porteus-v2.0-i486 bug reports

Posted: 05 Jul 2013, 14:59
by brokenman
You can load 'certain modules only' by putting them in the 'optional' folder and using the cheatcode load= (please read cheatcode.txt for more info). Using this method only the modules that you specify will be loaded from the optional folder.

Re: Porteus-v2.0-i486 bug reports

Posted: 05 Jul 2013, 20:46
by Rava
brokenman wrote:You can load 'certain modules only' by putting them in the 'optional' folder and using the cheatcode load= (please read cheatcode.txt for more info). Using this method only the modules that you specify will be loaded from the optional folder.
But I cannot leave my XFCe setting / base folder as it is and not load some of them while loading KDE ones instead, right?

Re: Porteus-v2.0-i486 bug reports

Posted: 05 Jul 2013, 21:28
by Ahau
I would do it like this:

Code: Select all

noload=xfce load=kde norootcopy
all without 'changes=' and 'xfce' so you're not saving changes and it doesn't try to load xfce. That should give you a clean testing platform and nothing will stick around after you reboot. You can drop 'noload=xfce' from these instructions if you still want to have mousepad around to test in. Does this happen if you use geany instead?

Re: Porteus-v2.0-i486 bug reports

Posted: 09 Jul 2013, 20:35
by Rava
So far I was not able to try out a clean KDE, but a new bug appeared (same PC, actually):

On my Samsung laptop I use a external firewire DVD burner, that one used to work okay, but for some reason now fails with 2.0 i586 (The machine is not able to run x86-64 stuff...) When I recall correct, it used to work with 2.0 in the past...

I get this info when plugging in the DVD burner:

Code: Select all

porteus kernel: [112609.924641] firewire_core 0000:09:09.1: created device fw1: GUID 0000f041e000427f, S400
porteus kernel: [112609.924657] firewire_core 0000:09:09.1: phy config: new root=ffc0, gap_count=5

Code: Select all

root@porteus:/mnt# cdrecord -scanbus
Cdrecord-ProDVD-ProBD-Clone 3.01a08 (i686-pc-linux-gnu) Copyright (C) 1995-2012 Joerg Schilling
cdrecord: No such file or directory. Cannot open '/dev/pg*'. Cannot open or use SCSI driver.
cdrecord: For possible targets try 'cdrecord -scanbus'.
cdrecord: For possible transport specifiers try 'cdrecord dev=help'.
Using /dev/fw1 results in this error:

Code: Select all

root@porteus:/iso/porteus/2.1rc2# cdrecord speed=10 dev=/dev/fw1 -tao -v driveropts=burnfree Porteus-XFCE-v2.1rc2-x86_64.iso
Cdrecord-ProDVD-ProBD-Clone 3.01a08 (i686-pc-linux-gnu) Copyright (C) 1995-2012 Joerg Schilling
TOC Type: 1 = CD-ROM
scsidev: '/dev/fw1'
devname: '/dev/fw1'
scsibus: -2 target: -2 lun: -2
Warning: Open by 'devname' is unintentional and not supported.
cdrecord: Inappropriate ioctl for device. Cannot open or use SCSI driver.
cdrecord: For possible targets try 'cdrecord -scanbus'.
cdrecord: For possible transport specifiers try 'cdrecord dev=help'.
Trying pburn is to no prevail, it tells me it choose /dev/cdrom as device, and giving me the list to choose the correct device from, but but the list is empty. :(
And sure, /dev/cdrom is non-existent.

Any ideas?

Re: Porteus-v2.0-i486 bug reports

Posted: 09 Jul 2013, 21:46
by fanthom
hi Rava,

please run 'lsmod' and check ig 'sg' driver is loaded. if not then please load it manually with 'modprobe sg' command. this driver is responsible for SCSI stuff including external optical writers so hopefully can solve your issue.
let me know if this is the case so i'll compile it directly into kernel.

thanks

Re: Porteus-v2.0-i486 bug reports

Posted: 10 Jul 2013, 02:05
by Rava
@fanthom
It was not loaded, but "modprobe sg" exits with code 1, so loading sg was not successfully. Any ideas why that can be?

Re: Porteus-v2.0-i486 bug reports

Posted: 10 Jul 2013, 14:51
by brokenman
Please try with kernel from latest rc2 if possible. This module loads okay for me.

Re: Porteus-v2.0-i486 bug reports

Posted: 10 Jul 2013, 16:39
by fanthom
"modprobe sg" exits with code 1
that's weird - did you mix up the kernel versions by any chance?
please show me what 'dmesg' says about this error (you could also upgrade to kernel from rc2 as suggested by brokenman).